The Fontes Anglo-Saxonici Database: The Stand-Alone Version

Malcolm Godden (database director, Fontes project) and Rohini Jayatilaka (database editor, Fontes project)

The stand-alone version of the Fontes Anglo-Saxonici database, a register of written sources used by Anglo-Saxon authors, was published in September 2002. We are grateful to the editor of OEN for an opportunity to describe the processes involved, our intentions in developing this version, and our plans for the future of the project.

Planning and Development

The Fontes Anglo-Saxonici database, which had been in progress since the late 1980s, was made publicly available in August 1999, in the form of an interactive web page on the project website (http://fontes.english.ox.ac.uk/). At the time it held some 16,000 records, covering 860 Anglo-Saxon texts and 850 sources and analogues. From that point scholars were able to interrogate the database online, asking to see the full source-details of every Anglo-Saxon text so far covered by the project, or a full listing of all the Anglo-Saxon texts which used a particular source, from Abbo's Passio Eadmundi to Vergil's Aeneid, and then to explore specific details of passages.

As a scholarly resource the site has proved extremely popular. We have registered an average of 500 hits a month on the site, on a regular basis in the three years and more since the launch. At the same time the database itself has been growing, thanks to the efforts of our own team and of scholars round the world. It now contains over 29,000 records analysing in detail the source-relationships of around 1,150 Anglo-Saxon texts (over 500 written in Old English and over 600 in Latin) and identifying the use of over 1000 sources and analogues.

The web-based system has proved remarkably robust over the period. The database itself is held on a server maintained by Oxford University Library Services, and updates are transferred regularly by Rohini from the Fontes office in the English Faculty, using ftp. "Down" periods have been rare, and service has been quickly restored by the OULS team. The Active Server Page queries developed by Paul Groves of the Humanities Computing Development Team in Oxford in 1999 to allow users to question the underlying database have continued to work well, and we have been able to make a few small adjustments since. Of the many occasions on which the Fontes team have demonstrated the capabilities of the site, perhaps the highlight has been the session at ISAS in Helsinki in 2001, when they were able to move seamlessly and in real time from the online version of the Old English corpus created by the Dictionary of Old English to the online version of the Fontes database and on to the online version of the Patrologia Latina, and so demonstrate the potentials for cross-website analyses.

Even so, the Fontes team have from the outset been aware of the inevitable limitations of web-based access to the database. Firstly, not all Anglo-Saxon scholars (despite their famous adaptability to new technology) had good access to the internet at all times, day and night, when they might feel an urgent need to consult the database, and for those dependent on modems and telephone lines (and worse still a telephone billing system less friendly than that familiar in the U.S.) prolonged consultation was not easy. More importantly, the nature of website systems and the limitations of the internet meant that consultation inevitably took the form of tightly focused questions of a set kind eliciting relatively small batches of information at a time, which could be easily downloaded to the user. You could ask for a list of the works of Vergil covered in the database; then you could pick one and ask for a list of Anglo-Saxon texts that used it; then you could pick one of those and ask for a list of passages in that text which used the Vergil poem; and then you could pick one of those passages and ask for the precise references and details; but it could be exasperating if what you really wanted to do was see a good range of material on screen and scroll up and down to get a sense of the range and pattern of use. And fundamentally, users couldn't get at the database itself and ask their own complicated or individual questions, or generate reports from it.

It was for these reasons that the Fontes team started looking at alternatives to the web version of the database from the moment it was launched. At first the task looked straightforward. The database used the Microsoft Access 97 program, and it would be easy to copy it on to a series of CD-ROMs and distribute these to scholars round the world. But this would only be usable by those who had MS Access installed on their own computer systems, or could readily convert it to a different application, and even then it would require a fair amount of familiarity with database programming and with the darker corners of the Fontes database to use the material fully. An alternative we considered was to copy the ASP queries used on the website and provide them on a CD-ROM along with the database itself. But this was to reproduce some of the limitations of web systems which we were trying to avoid. The eventual solution on which we settled was threefold: to develop a series of forms and queries within MS Access which could be packaged into a user-friendly front end of the database, facilitating access for those who were not IT buffs themselves; to adapt a "runtime" version of MS Access [1] as part of the package, so that it would operate on PCs that did not have MS Access itself installed; and to include as well a copy of the basic database so that it could be interrogated and manipulated by users who were comfortable with database applications in any way they chose.

A trawl of resources and experts available within the university failed at first to reveal anyone with the necessary expertise, so Rohini began developing the "forms" herself, alongside her other work of processing and creating entries for the database. Fortunately, after she had made a fair amount of progress and established the outlines of what was needed, we hit lucky and found the brilliant and resourceful David Miles, who was just taking early retirement from OUCS and planning to work freelance as a database designer. David quickly got on top of the project and developed for us a very attractive front-end, providing facilities that we hadn't thought of asking for. Perhaps the most interesting and useful (at least, we find it so) is what David called Author Reference Summaries, described below. David also developed the runtime application for us. Testing and refining the first version was an intensive business, as we had to try it out on a variety of machines and systems and imagine all the different queries users might want to make, or the advised and ill-advised routes they might take. Once we were fairly confident we burned the package (database, front-end application, information files in html, and runtime Access application) on to several CD-ROMs and asked the Fontes committee members to try it out, and once they gave positive responses we went for production. A local firm, Databiz, pressed 500 copies for us, with the Fontes name printed on each, and packaged them in CShells, a robust and cheap alternative to the traditional jewel-cases, while the university printer printed an accompanying booklet.

Publication and Distribution

Distribution was at first a puzzle. We were anxious to make the package widely available, and as freely or cheaply as possible, and so chose not to use a commercial publisher. But we did not have the resources to distribute very large numbers of CDs to individual users. We came up with a variety of alternative methods. Firstly, we have made it available as a download from the project website; scholars can use this to copy the whole (zipped) package directly onto their own PCs. Secondly, we are mailing CDs and booklets, as long as disks and postage funds last, to those who contact us and cannot use the download facility or prefer to have the package on CD-ROM. Thirdly, we have deposited copies with the Oxford Text Archive, which has agreed to make it available on demand both as a download and as a CD-ROM. The last route was particularly important to us because the CD-ROM version of the database is itself a scholarly publication which will be cited in research publications, and it was therefore crucial that the package should be available into the future from an established institution, which would continue in existence in some form after the Fontes team had disbanded.

Why CD-ROM?

The answers to this question are perhaps already evident, but since the question was repeatedly asked at the outset, and the use of this medium was still generating comment and complaint at the end, we think we need to explain. Initially, the answer was that the web can do some things remarkably well, but with material like ours it cannot give full access and cannot do all we would like the Fontes database to do for scholarship. Secondly, websites are ephemeral, or at least transient, while scholarship is for ever; the project team cannot undertake to maintain the web version of the database indefinitely, and we do not wish to see the immense amount of scholarship and scholarly resource invested in it lost. But in the end, the answer is that this is not "a CD-ROM version," though we have to admit that that is what we have called it to distinguish it clearly from the web version. It is in fact a stand-alone version, designed to be installed on an individual scholar's PC wherever it is and used however the scholar wishes to use it. The CD-ROM is one of the means by which we distribute this version, and users can leave the data on the CD if they wish (if they have little room on their own hard-disk for instance), but it is not a necessary feature. We could of course have distributed it on DVD instead or as well, but for most potential users CD-ROM is still the more common facility. The package can also be networked from a server within an institution, simply by installing it.

The real problem here (we thought) was libraries, which don't like buying or housing CD-ROM material, still less making it accessible, and press hard for web versions with annual subscriptions instead of one-off CD-ROM purchases. This seems to us understandable but short-sighted (the last thing you would expect to accuse libraries of). The web version of Patrologia Latina, maintained by annual subscription, is vulnerable to the next cut in library funding or the next financial crisis of a publisher; the CD-ROM version, once bought, is in the library for ever, and a resourceful IT center can ensure that as technology changes the material can be adapted. We were very pleased to find that within days of distributing the first copies of the CD-ROM we were receiving requests from colleagues for additional copies for their libraries, which were keen to have it.

What Can the Stand-Alone Fontes Application Do?

The opening screen offers three options:

Information takes you to a variety of documents giving you information about the project and its use.

Fontes on the Web takes you into the database on the web, using your internet browser and connection.

Fontes Database takes you into the stand-alone version of the database.

If you pursue the last of these, you enter the database itself. The default page is Anglo-Saxon Texts and shows a list of the Anglo-Saxon authors and the 1,152 works by them which are included in the database, from Abbo of Fleury's Passio Sancti Eadmundi to Wulfstan of York's Homily 21 (yes, we know that Abbo was not an Anglo-Saxon author, but since he very probably wrote the text in England and may have used sources available in England, the text qualifies for the Fontes register).

If your interest is in, say, the texts drawn on by Bede for his prose Life of St. Cuthbert, whether as major sources or for isolated phrases and quotations, then you can scroll down to Beda, Vita S. Cuthberti (prose), or use the search boxes at the top of the screen, and then click on the button Show Sources for Selected A-S Text, and you will get a full list of the 61 records we have for that text, from the quotation of Psalm 26:13 at 146:4-5 (the reference is to page and line in Colgrave's edition of the Vita) down to the phrase lifted from Gregory the Great's Dialogi at 202.8. Users familiar with the web version should immediately notice the range of information instantly available in the stand-alone version. If your fancy prompts you to ask how other Anglo-Saxon writers might have used Psalm 26.13, then you can highlight that record and click on the button Source Text, which will immediately take you to a list of records of Anglo-Saxon texts using the psalms, and in particular the evidence of three Latin charters which also use that precise verse.

Anonymous texts are of course trickier to track down than a work by Bede, and an exasperating number of Anglo-Saxon texts are still anonymous. We have tried to help by dividing "Anonymous" as author into two major groups, Anon (OE) and Anon (Lat.), for English and Latin texts, and further separated Anon (OE Martyrology) to isolate the nearly 300 texts which form the Old English Martyrology. An additional help is the new device for searching by any part of the title. Typing the name "Andrew" into the Title box at the top of the screen will take you instantly to two relevant Anglo-Saxon texts, the entry on St. Andrew in the OE Martyrology and the homily on St. Andrew which occurs in the Blickling collection.

The kinds of information about texts and sources which the database has always contained are now more immediately available on screen from the outset. These include references to the transmission history of sources, relevant bibliography, comments by the contributor and (for saints' lives) the reference number in Bibliotheca Hagiographica Latina. The suggested relationship between a particular passage and a particular "source-passage" (certain source, probable one, possible one, one of several sources for the passage, mere analogue in default of sources) is still signalled by a mysterious looking "sigil" such as M3o, but a new button called Explain Sigla will produce a dialogue box that explains, for instance, that M3o against a detail from the Martyrology account of Andrew means that the passage cited from Isidore is no more than one possible source, and the user can see the other possibilities on the screen.

If your interest is in earlier texts which may have been used by Anglo-Saxon writers, then you call up the

Source Texts screen by clicking on Show Source Texts, and get the list of source authors and the 880 works by them which figure as sources or analogues in the register. Scrolling down to Vergil will show you the three works of his which are used by Anglo-Saxon authors, the Aeneid, the Eclogues and the Georgics. Selecting the Aeneid will give you a list of the 243 passages from this poem which the database records, from Book 1 line 142 to Book 9 line 769 or, using the different way of ordering them that the application allows (alphabetically by Anglo-Saxon author and text), from the use by Abbo of Fleury at 1.209 in his Passio Eadmundi to the use by Wulfstan Cantor at 9.340 in his metrical life of St. Swithun. From any one of these records you can in the new application move instantly to the full set of records for the Anglo-Saxon text in question, giving you, for instance, the record for 9.340 of Wulfstan's Vita metrica along with the adjacent records for this text. If alternatively you want to ascertain, say, which of the Old Testament prophets was most used by Anglo-Saxon writers, then picking on Biblia Sacra (the project name for the corporate author of the Bible) and scrolling down the different books with an eye to the number of records for each at the bottom of the screen will rapidly reveal that the answer, by a long way, is Isaiah, with 258 hits.

A new feature which we are very pleased with is the Author Reference Summaries, a page which assembles all the Anglo-Saxon authors on the left of the screen and all the Source Authors on the right. Highlighting an Anglo-Saxon author on the left produces on the right a list of the Source Authors that his work uses; highlighting a Source Author generates on the left a list of all the Anglo-Saxon authors who use this author's work already described above. We have also included the various bibliographies which are available in the web version, but here we have merged them into one full list, covering publications of Anglo-Saxon texts, editions of source-texts and works which provide identifications of sources. We have added several ways in which to search the bibliographies and to sort the resulting records of a search.

All this describes what you can do using the new front-end application as a route into the Fontes database. But what if you have MS Access or a similar database program and are willing to develop your own queries? Now you can just make another copy of the Fontes database which is in our package on your hard disk, untick the "read-only" box and open it in Access. Then you can, for instance, write a simple query asking for all the Biblical passages quoted in Alfredian works, or design and print a report giving you hard copy of the sources of Bede's metrical Life of St. Cuthbert.

The Future of the Fontes Register

The project has reached one of its major targets, the publication of a flexible and widely-available version of the database, but has not yet finished its work. There are still texts which ought to be covered, especially in Latin, and scholars round the world who have promised to send us records for those texts based on their own research. As more material comes in, we will be editing it and adding it to the database, and publishing it in the web version of the database. But for those who are using the stand-alone version, we will also be placing up-dated versions of the database on our website in zipped form for downloading, from time to time. Users who have the stand-alone package on their PC can simply download the new copy of the whole database, and substitute it for the original data file (FontesData.mdb). Users who are relying on the CD-ROM for the data, for instance those using library copies, will not be able to do this, though it is feasible to burn a new CD-ROM and add the new data. We hope to announce

the release of any updates on various electronic mail lists, such as the ISAS list and ANSAXNET. In the longer term (three to five years), we hope to issue a second edition of the stand-alone version, containing updated data and perhaps enhancements of the application. (And yes, we will be looking hard at the possibility of producing a stand-alone version for Macs.)

Publication and Permanence

One of the questions that has exercised the project team is how to ensure that the material remains available into the future. The records for particular texts contain scholarly discoveries and conclusions that are frequently not published in any other form, and we have from the outset maintained the position that each set of records was itself a publication which should be noted in bibliographies (early contributions were reported in the Anglo-Saxon England bibliography, more recent ones in the project report published each year in the Old English Newsletter). In addition, the database as a whole provides information of a research-based kind that is not available from any individual set of records or from hard-copy publication (it is now a matter of some wry amusement that the earliest meetings of the project, in the 1980s, included detailed discussions of the kind of book-form publication that we would be expecting as the culmination of the project). And although we have always sought not to control or restrict access to the database, we have asked, and do ask, users to acknowledge its use in their publications and, where appropriate, the individual contributors whose work they have used.

It follows that it has been incumbent on us to ensure that any potential user, perhaps one following up a reference in a publication, could get access to the database, both now and into the future (and the future can be very long in our subject). That is a difficult proposition with a register that exists only in electronic form. The web version is universally available to anyone with a browser and internet connection, but as we have noted we cannot guarantee to maintain the web access indefinitely, dependent as it is on institutional support and individual responsibility. That, indeed, was one of the reasons for exploring the possibility of CD-ROM publication in the first place. Publication and distribution on CD-ROM does ensure that there are physical copies of the register available around the world, now and for the future. But while we can make sure by energetic distribution, and such publications as this, that current scholars are aware of the availability of the database and can obtain copies, this does not help with future scholars or with those outside the field of Anglo-Saxon studies who might come to need to consult our work. The usual answer, with books and articles, is of course that copies should be available in scholarly libraries, but their record of acquisition is far from impressive with electronic publications. Tentative researches revealed that some of the most important recent publications in our field were not being acquired by libraries (the Bodleian does not have Kevin Kiernan's Electronic Beowulf and now cannot obtain a copy). As noted above, some libraries proved less than enthusiastic about acquiring CD-ROM material, and it was far from clear that those which did, or would, would actually make it accessible, particularly with a product like the Fontes application which needed to be installed on a PC rather than just read with a browser. Nor, of course, can we be sure that either the software or the hardware that the application requires will be available in libraries or on individuals' PCs into the indefinite future -- or rather, we can be sure that it won't be.

A partial solution, we found, was with the Oxford Text Archive, which operates as an arm of the Arts and Humanities Data Service in the UK and has an established role in maintaining archives of electronic data produced by publicly funded scholarship. The OTA has agreed to keep stock of the CD-ROM version and to maintain a copy on its own server, and to provide and distribute copies into the future, and it also has a general interest in maintaining access to such material long-term. We can see the possibility of converting the database, once it is finalised, into a simpler all-purpose form, but still on disk, which can more easily be adapted to future database applications and different media. But it is impossible to envisage the kind of scholarship it incorporates being usefully converted into hard-copy format. Clearly if humanities scholarship is to continue to publish original research in electronic form, we need to ensure that there are more places where access can be provided into the future.

In the meantime, the Fontes project continues under the chairmanship of Don Scragg at Manchester University, and more contributions to the database will be welcome (not least from those who committed themselves some years ago, and who are STILL on our list). Enquiries about Old English texts should be sent to Dr Susan Irvine at University College London and about Anglo-Latin texts to Dr Rosalind Love at Cambridge University. And enquiries about the database itself, or requests for copies, should be sent to the abovesigned at the Fontes Anglo-Saxonici Project, English Faculty, University of Oxford, St. Cross Building, Manor Road, Oxford OX1 3UQ, United Kingdom.

Notes

[1] The runtime version is a facility provided by the MS Access 2000 Professional Developer Edition.